Riemann sum
Very similar to this: Riemann Sum (univariate calculus) .
Intuition define:
Now give you a partition P P P :
P : I = ⋃ j = 1 m I j P: I=\bigcup_{j=1}^{m} I_{j} P : I = j = 1 ⋃ m I j of this interval with distinguished points ξ = { ξ 1 , ⋯ , ξ k } \xi=\left\{\xi_{1}, \cdots, \xi_{k}\right\} ξ = { ξ 1 , ⋯ , ξ k } .
We defined Riemann sum as:
σ ( f , P , ξ ) : = ∑ i = 1 k f ( ξ i ) ∣ I i ∣ \sigma(f, P, \xi):=\sum_{i=1}^{k} f\left(\xi_{i}\right)\left|I_{i}\right| σ ( f , P , ξ ) := i = 1 ∑ k f ( ξ i ) ∣ I i ∣ If lim ∣ P ∣ → 0 σ ( f , P , ξ ) \lim _{|P| \rightarrow 0} \sigma(f, P, \xi) lim ∣ P ∣ → 0 σ ( f , P , ξ ) exist and finite, we often denoted it as ∫ I f ( x ) d x \int_I f(x)\mathrm d x ∫ I f ( x ) d x or :
∫ I f ( x 1 , ⋯ , x n ) d x 1 ⋯ d x n , ∫ ⋯ ∫ f ( x 1 , ⋯ , x n ) d x 1 ⋯ d x n \int_{I} f\left(x^{1}, \cdots, x^{n}\right) \mathrm{d} x^{1} \cdots \mathrm{d} x^{n}, \quad \int \cdots \int f\left(x^{1}, \cdots, x^{n}\right) \mathrm{d} x^{1} \cdots \mathrm{d} x^{n} ∫ I f ( x 1 , ⋯ , x n ) d x 1 ⋯ d x n , ∫ ⋯ ∫ f ( x 1 , ⋯ , x n ) d x 1 ⋯ d x n Fubini’s Theorem
Let X × Y X \times Y X × Y be an interval in R m + n \mathbb{R}^{m+n} R m + n , which is the direct product of intervals X ⊂ R m X \subset \mathbb{R}^{m} X ⊂ R m and Y ⊂ R n Y \subset \mathbb R^{n} Y ⊂ R n . If the function f : X × Y → R f: X \times Y \rightarrow \mathbb{R} f : X × Y → R is integrable over X × Y X \times Y X × Y , then all three of the integrals
∫ X × Y f ( x , y ) d x d y , ∫ X d x ∫ Y f ( x , y ) d y , ∫ Y d y ∫ X f ( x , y ) d x \int_{X \times Y} f(x, y) \mathrm{d} x \mathrm{d} y, \quad \int_{X} \mathrm{d} x \int_{Y} f(x, y) \mathrm{d} y, \quad \int_{Y} \mathrm{d} y \int_{X} f(x, y) \mathrm{d} x ∫ X × Y f ( x , y ) d x d y , ∫ X d x ∫ Y f ( x , y ) d y , ∫ Y d y ∫ X f ( x , y ) d x exist and are equal.
Change of Variable in a Multiple Integral
If φ : D t → D x \varphi: D_{t} \rightarrow D_{x} φ : D t → D x is a diffeomorphism of a bounded open set D t ⊂ R n D_{t} \subset \mathbb{R}^{n} D t ⊂ R n onto a set D x = φ ( D t ) ⊂ R n D_{x}=\varphi\left(D_{t}\right) \subset \mathbb{R}^{n} D x = φ ( D t ) ⊂ R n of the same type, f ∈ R ( D x ) , f \in \mathcal{R}\left(D_{x}\right), f ∈ R ( D x ) , and supp f \operatorname{supp} f supp f is a compact subset of D x D_{x} D x , then f ∘ φ ∣ det φ ′ ∣ ∈ R ( D t ) f \circ \varphi\left|\operatorname{det} \varphi^{\prime}\right| \in \mathcal{R}\left(D_{t}\right) f ∘ φ ∣ det φ ′ ∣ ∈ R ( D t ) , and the following formula holds:
simplified Chinese: 设 φ : D t → D x \varphi: D_{t} \rightarrow D_{x} φ : D t → D x 为 R n \mathbb{R}^{n} R n 中的有界开集之间的微分同胚,f f f 为 定义在 D x D_{x} D x 上的函数,满足条件 supp f ⊂ D x . \operatorname{supp} f \subset D{x} . supp f ⊂ D x . 则 f ∈ R ( D x ) f \in \mathcal{R}\left(D{x}\right) f ∈ R ( D x ) 当且仅当 f ∘ φ ⋅ ∣ det φ ′ ∣ ∈ R ( D t ) f \circ \varphi \cdot\left|\operatorname{det} \varphi^{\prime}\right| \in \mathcal{R}\left(D_{t}\right) f ∘ φ ⋅ ∣ det φ ′ ∣ ∈ R ( D t ) , 并且如果 f ∈ R ( D x ) f \in \mathcal{R}\left(D_{x}\right) f ∈ R ( D x ) ,则成立如下的变量代换公式:
∫ D x f ( x ) d x = ∫ D t f ∘ φ ⋅ ∣ det φ ′ ∣ d t \int_{D_{x}} f(x) \mathrm{d} x=\int_{D_{t}} f \circ \varphi \cdot\left|\operatorname{det} \varphi^{\prime}\right| \mathrm{d} t ∫ D x f ( x ) d x = ∫ D t f ∘ φ ⋅ ∣ det φ ′ ∣ d t Example
(1) Polar coordinate transformation
In general polar coordinates ( r , θ 1 , ⋯ , θ n − 1 ) (r,\theta_1,\cdots,\theta_{n−1}) ( r , θ 1 , ⋯ , θ n − 1 ) in R n \mathbb R^n R n are introduced via the relations:
{ x 1 = r cos θ 1 x 2 = r sin θ 1 cos θ 2 ⋮ x n − 1 = r sin θ 1 sin θ 2 ⋅ … sin θ n − 2 cos θ n − 1 x n = r sin θ 1 sin θ 2 ⋅ … ⋅ sin θ n − 1 sin θ n − 1 \left\{\begin{array}{l}
x^{1}\quad=r \cos \theta_{1} \\
x^{2}\quad=r \sin \theta_{1} \cos \theta_{2} \\
\qquad\vdots \\
x^{n-1}=r \sin \theta_{1} \sin \theta_{2} \cdot \ldots \sin \theta_{n-2} \cos \theta_{n-1} \\
x^{n}\quad=r \sin \theta_{1} \sin \theta_{2} \cdot \ldots \cdot \sin \theta_{n-1} \sin \theta_{n-1}
\end{array}\right. ⎩ ⎨ ⎧ x 1 = r cos θ 1 x 2 = r sin θ 1 cos θ 2 ⋮ x n − 1 = r sin θ 1 sin θ 2 ⋅ … sin θ n − 2 cos θ n − 1 x n = r sin θ 1 sin θ 2 ⋅ … ⋅ sin θ n − 1 sin θ n − 1 because of the orthogonality, we have:
∣ det ψ ′ ∣ 2 = det ψ ′ ⋅ det ( ψ ′ ) T = ∥ ∂ x ∂ r ∥ ∥ ∂ x ∂ θ 1 ∥ ⋯ ∥ ∂ x ∂ θ n − 1 ∥ = ( r n − 1 sin n − 2 θ 1 sin n − 3 θ 2 … sin θ n − 2 ) 2 \begin{aligned}
\left|\operatorname{det} \psi^{\prime}\right|^{2} &=\operatorname{det} \psi^{\prime} \cdot \operatorname{det}\left(\psi^{\prime}\right)^{T}=\left\|\frac{\partial x}{\partial r}\right\|\left\|\frac{\partial x}{\partial \theta_{1}}\right\| \cdots\left\|\frac{\partial x}{\partial \theta_{n-1}}\right\| \\
&=\left(r^{n-1} \sin ^{n-2} \theta_{1} \sin ^{n-3} \theta_{2} \ldots \sin \theta_{n-2}\right)^{2}
\end{aligned} ∣ det ψ ′ ∣ 2 = det ψ ′ ⋅ det ( ψ ′ ) T = ∂ r ∂ x ∂ θ 1 ∂ x ⋯ ∂ θ n − 1 ∂ x = ( r n − 1 sin n − 2 θ 1 sin n − 3 θ 2 … sin θ n − 2 ) 2 so we recall the Jacobian :
J = r n − 1 sin n − 2 θ 1 sin n − 3 θ 2 ⋅ … ⋅ sin θ n − 2 J=r^{n-1} \sin ^{n-2} \theta_{1} \sin ^{n-3} \theta_{2} \cdot \ldots \cdot \sin \theta_{n-2} J = r n − 1 sin n − 2 θ 1 sin n − 3 θ 2 ⋅ … ⋅ sin θ n − 2 (2) The volume of
n − n- n − dimensional sphere
Denote the sphere is B n ( r ) B_n(r) B n ( r ) We can easily have the recurrence formula:
V n ( r ) = ∫ − r r d x ∫ B n − 1 ( r 2 − x 2 ) d x 2 ⋯ d x n = ∫ − r r V n − 1 ( r 2 − x 2 ) d x V_{n}(r)=\int_{-r}^{r} \mathrm{d} x \int_{B_{n-1}\left(\sqrt{r^{2}-x^{2}}\right)} \mathrm{d} x^{2} \cdots \mathrm{d} x^{n}=\int_{-r}^{r} V_{n-1}\left(\sqrt{r^{2}-x^{2}}\right) \mathrm{d} x V n ( r ) = ∫ − r r d x ∫ B n − 1 ( r 2 − x 2 ) d x 2 ⋯ d x n = ∫ − r r V n − 1 ( r 2 − x 2 ) d x It's trivial that V 1 ( r ) = 2 r V_1(r)=2r V 1 ( r ) = 2 r , V 2 ( r ) = π r 2 V_2(r)=\pi r^2 V 2 ( r ) = π r 2 , assume that V n ( r ) = c n r n V_{n}(r)=c_{n} r^{n} V n ( r ) = c n r n , so
V n + 1 ( r ) = ∫ − r r c n ( r 2 − x 2 ) n 2 d x = c n ∫ − π 2 π 2 r n cos n φ ⋅ r cos φ d φ = c n ∫ − π 2 π 2 r n + 1 cos n + 1 φ d φ = c n + 1 r n + 1 \begin{aligned}
V_{n+1}(r) &=\int_{-r}^{r} c_{n}\left(r^{2}-x^{2}\right)^{\frac{n}{2}} \mathrm{d} x=c_{n} \int_{-\frac{\pi}{2}}^{\frac{\pi}{2}} r^{n} \cos ^{n} \varphi \cdot r \cos \varphi \mathrm{d} \varphi \\
&=c_{n} \int_{-\frac{\pi}{2}}^{\frac{\pi}{2}} r^{n+1} \cos ^{n+1} \varphi \mathrm{d} \varphi=c_{n+1} r^{n+1}
\end{aligned} V n + 1 ( r ) = ∫ − r r c n ( r 2 − x 2 ) 2 n d x = c n ∫ − 2 π 2 π r n cos n φ ⋅ r cos φ d φ = c n ∫ − 2 π 2 π r n + 1 cos n + 1 φ d φ = c n + 1 r n + 1 with
c n + 1 = c n ∫ − π 2 π 2 cos n + 1 φ d φ c_{n+1}=c_{n} \int_{-\frac{\pi}{2}}^{\frac{\pi}{2}} \cos ^{n+1} \varphi \mathrm{d} \varphi c n + 1 = c n ∫ − 2 π 2 π cos n + 1 φ d φ so that,
V n ( r ) = { 2 ( 2 π ) k ( 2 k + 1 ) ! ! r 2 k + 1 , n = 2 k + 1 ( 2 π ) k ( 2 k ) ! ! r 2 k , n = 2 k V_{n}(r)=\left\{\begin{array}{ll}
2 \frac{(2 \pi)^{k}}{(2 k+1) ! !} r^{2 k+1}, & n=2 k+1 \\
\frac{(2 \pi)^{k}}{(2 k) ! !} r^{2 k}, & n=2 k
\end{array}\right. V n ( r ) = { 2 ( 2 k + 1 )!! ( 2 π ) k r 2 k + 1 , ( 2 k )!! ( 2 π ) k r 2 k , n = 2 k + 1 n = 2 k ■ \blacksquare ■